CODEWITHSUNDEEP
rggplot2
jwillis0720
jwillis0720

Reputation: 4477

ggplot2 - area fill is incorrectly plotting the area

I thought this was a problem with the ordering of how I was placing each aestetic. Here is the relevant code.

enter image description here

Sort of like a normal distribution. Going to melt it down and plot each distribution.

 df <- as.data.frame(r_data_frame)
 names(df) <- c('Length','Unique','Donor Unique','All')
 library(ggplot2)
 library(reshape2)
 ndf <- melt(df,id.vars='Length')
 print(ndf)
 graph <- ggplot(data=ndf) + geom_area(aes(Length,value,fill=variable)) +
     theme_bw()+
     xlim(1,42)+
     geom_hline(yintercept=2369802) +
     geom_hline(yintercept=2469225,color='red')+
     geom_vline(xintercept=15)

 plot(graph)

The melted data frame is probably better found in this gist

Here is the output

enter image description here

The problem I'm having is that the shaded areas don't seem to be plotted correctly. I plotted the horizontal lines for both "Donor Unique" and "Unique" of where they should be according to the data frame. "Unique" is right as it is right at its y value. However "Donor Unique" is way to high. They should barely overlap. If you look at the values What in the world is going on? Should I be using something else besides geom_area?

Upvotes: 0

Views: 514

Answers (1)

jlhoward
jlhoward

Reputation: 59355

You need to use position="dodge" in the call to geom_area(...).

Many of the geom_* functions in ggplot, including geom_area(...), geom_histogram(...), and geom_bar(...), have a position parameter. The default is (usually...) "stacked", which creates geometries that are additive. To display the geometries with independent y-values, use position="dodge".

Upvotes: 1

Related Questions